final_version.pdf (1.22 MB)
Download file

Applying lightweight soft error mitigation techniques to embedded mixed precision deep neural networks

Download (1.22 MB)
journal contribution
posted on 15.11.2021, 09:55 authored by Geancarlo Abich, Jonas Gava, Rafael Garibotti, Ricardo Reis, Luciano OstLuciano Ost
Deep neural networks (DNNs) are being incorporated in resource-constrained IoT devices, which typically rely on reduced memory footprint and low-performance processors. While DNNs' precision and performance can vary and are essential, it is also vital to deploy trained models that provide high reliability at low cost. To achieve an unyielding reliability and safety level, it is imperative to provide electronic computing systems with appropriate mechanisms to tackle soft errors. This paper, therefore, investigates the relationship between soft errors and model accuracy. In this regard, an extensive soft error assessment of the MobileNet model is conducted considering precision bitwidth variations (2, 4, and 8 bits) running on an Arm Cortex-M processor. In addition, this work promotes the use of a register allocation technique (RAT) that allocates the critical DNN function/layer to a pool of specific general-purpose processor registers. Results obtained from more than 4.5 million fault injections show that RAT gives the best relative performance, memory utilization, and soft error reliability trade-offs w.r.t. a more traditional replication-based approach. Results also show that the MobileNet soft error reliability varies depending on the precision bitwidth of its convolutional layers.

History

School

  • Mechanical, Electrical and Manufacturing Engineering

Published in

IEEE Transactions on Circuits and Systems I: Regular Papers

Volume

68

Issue

11

Pages

4772 - 4782

Publisher

Institute of Electrical and Electronics Engineers (IEEE)

Version

AM (Accepted Manuscript)

Rights holder

© IEEE

Publisher statement

Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.

Acceptance date

13/07/2021

Publication date

2021-07-26

Copyright date

2021

ISSN

1549-8328

eISSN

1558-0806

Language

en

Depositor

Dr Luciano Ost. Deposit date: 12 November 2021